Diabetes Prediction Through Linkage of Causal Discovery and Inference Model with Machine Learning Models
Problem
Predicting diabetes and identifying causal risk factors from patient health data, to understand contributors to a growing global disease.
Why it matters
Diabetes is increasing worldwide and brings serious complications such as cardiovascular disease.
What the AI did
Machine learning models (logistic regression, deep learning, gradient boosting, decision trees) predicted diabetes, and LiNGAM causal discovery inferred direct causal contributors among variables.
Results
Logistic regression 84.84% and deep learning 84.83% accuracy; causal factors identified: physical activity, difficulty walking, heavy drinking
Impact
Enables diabetes prediction and causal factor identification on a public dataset; clinical adoption not evidenced.
Record
| Venue | Biomedicines |
|---|---|
| Published | 2025 |
| Work type | article |
| Authors | Minseong Noh, Yang Sok Kim |
| Citations | 10 (OpenAlex, at import time) |
| References | 15 |
| Model or method | logistic regression, deep learning, gradient boosting, decision trees; LiNGAM |
| Organization | not stated in source |
| Primary source | https://doi.org/10.3390/biomedicines13010124 |
| Open access | gold |
| Keywords | diabetes prediction, machine learning, causal inference, LiNGAM, logistic regression |
Importance rating
major — millions of lives, an industry or a field
Abstract states global diabetes population continues to increase with serious complications including cardiovascular disease; a major worldwide disease affecting millions.